Computational identification of host genomic biomarkers highlighting their functions, pathways and regulators that influence SARS-CoV-2 infections and drug repurposing

Abstract

The pandemic threat of COVID-19 has severely destroyed human life as well as the economy around the world. Although, the vaccination has reduced the outspread, but people are still suffering due to the unstable RNA sequence patterns of SARS-CoV-2 which demands supplementary drugs. To explore novel drug target proteins, in this study, a transcriptomics RNA-Seq data generated from SARS-CoV-2 infection and control samples were analyzed. We identified 109 differentially expressed genes (DEGs) that were utilized to identify 10 hub-genes/proteins (TLR2, USP53, GUCY1A2, SNRPD2, NEDD9, IGF2, CXCL2, KLF6, PAG1 and ZFP36) by the protein-protein interaction (PPI) network analysis. The GO functional and KEGG pathway enrichment analyses of hub-DEGs revealed some important functions and signaling pathways that are significantly associated with SARS-CoV-2 infections. The interaction network analysis identified 5 TFs proteins and 6 miRNAs as the key regulators of hub-DEGs. Considering 10 hub-proteins and 5 key TFs-proteins as drug target receptors, we performed their docking analysis with the SARS-CoV-2 3CL protease-guided top listed 90 FDA approved drugs. We found Torin-2, Rapamycin, Radotinib, Ivermectin, Thiostrepton, Tacrolimus and Daclatasvir as the top ranked seven candidate drugs. We investigated their resistance performance against the already published COVID-19 causing top-ranked 11 independent and 8 protonated receptor proteins by molecular docking analysis and found their strong binding affinities, which indicates that the proposed drugs are effective against the state-of-the-arts alternatives independent receptor proteins also. Finally, we investigated the stability of top three drugs (Torin-2, Rapamycin and Radotinib) by using 100 ns MD-based MM-PBSA simulations with the two top-ranked proposed receptors (TLR2, USP53) and independent receptors (IRF7, STAT1), and observed their stable performance. Therefore, the proposed drugs might play a vital role for the treatment against different variants of SARS-CoV-2 infections.

Figure 1

The pipeline of this study.

Figure 2

(a) Volcano plot of the DEGs by edgeR Method and (b) by DESeq2 method. (c) Venn diagram of two DEGs-sets identified by DESeq2 and edgeR to show the common and uncommon genes. (d) The scatter plot first two principal components (PCs) of DEGs to see their prognostic performance of the case vs control.

Figure 3

The PPI network of DEGs.

Figure 4

The top 10 significantly enriched KEGG pathways with the hub-DEGs/proteins. The associated hub-DEGs are displayed in the right side of each bar. The hub-DEGs with bold represents upregulated genes and others are downregulated genes.

Figure 5

The TFs and hub-DEGs interaction network, where the blue squared nodes represent the TFs and the red round nodes represent the hub-DEGs. The key TFs are represented by the larger squared nodes.

Figure 6

The miRNAs and hub-DEGs interaction network where the small green octagonal nodes stand for miRNAs and the round nodes with red color represents the hub-DEGs. The key miRNAs are represented by the larger highlighted octagonal shaped green colored nodes in the figure.

Figure 7

(a) The disease versus hub-DEGs interaction network finds the comorbidities (b) The multivariate survival curves of lung cancer patients based on hub-DEGs.

Figure 8

Molecular docking results computed with autodock vina. Red colors indicated the strong binding affinities between target proteins and drug agents, and green colors indicated their weak bindings. (a) Image of binding affinities based on the top 50 ordered drug agents out of 90 in X-axis and ordered 15 target proteins (proposed) in Y-axis. (b) Image of binding affinities based on the proposed ordered 7 candidate drugs in X-axis and ordered 11 independent receptors (already published) in Y-axis. (c) Image of binding affinities based on the proposed ordered 7 candidate drugs in X-axis and original & protonated (at pH-7)* receptors in Y-axis, where * indicates the protonated receptors.

Figure 9

The RMSD analysis results for the variations of moving and initial drug-target complexes with 100 ns based MD simulations. (a) represented the RMSD with proposed top ranked two receptors (TLR2, USP53) and (b) represented the RMSD with top ranked two independent receptors (IRF7, STAT1).

Figure 10

Binding free energy (in kJ mol⁻¹) of each snapshot was calculated through 100 ns MD-based MM-PBSA simulations (a) represented the binding free energy with proposed top ranked two receptors (TLR2, USP53) and (b) represented the binding free energy with top ranked two independent receptors (IRF7, STAT1).